US11249158B2 - Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS) - Google Patents

Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS) Download PDF

Info

Publication number
US11249158B2
US11249158B2 US16/324,193 US201716324193A US11249158B2 US 11249158 B2 US11249158 B2 US 11249158B2 US 201716324193 A US201716324193 A US 201716324193A US 11249158 B2 US11249158 B2 US 11249158B2
Authority
US
United States
Prior art keywords
scanner
mrs
data
target
scanners
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active, expires
Application number
US16/324,193
Other languages
English (en)
Other versions
US20200191891A1 (en
Inventor
John M. Irvine
Laura J. Mariano
Alexander P. Lin
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Charles Stark Draper Laboratory Inc
Brigham and Womens Hospital Inc
Original Assignee
Charles Stark Draper Laboratory Inc
Brigham and Womens Hospital Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Charles Stark Draper Laboratory Inc, Brigham and Womens Hospital Inc filed Critical Charles Stark Draper Laboratory Inc
Priority to US16/324,193 priority Critical patent/US11249158B2/en
Publication of US20200191891A1 publication Critical patent/US20200191891A1/en
Assigned to THE CHARLES STARK DRAPER LABORATORY, INC. reassignment THE CHARLES STARK DRAPER LABORATORY, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: IRVINE, JOHN M., MARIANO, LAURA J.
Assigned to THE BRIGHAM AND WOMEN'S HOSPITAL, INC. reassignment THE BRIGHAM AND WOMEN'S HOSPITAL, INC. ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: LIN, Alexander P.
Application granted granted Critical
Publication of US11249158B2 publication Critical patent/US11249158B2/en
Assigned to UNITED STATES GOVERNMENT reassignment UNITED STATES GOVERNMENT CONFIRMATORY LICENSE (SEE DOCUMENT FOR DETAILS). Assignors: CHARLES STARK DRAPER LABORATORY
Assigned to UNITED STATES GOVERNMENT reassignment UNITED STATES GOVERNMENT CONFIRMATORY LICENSE (SEE DOCUMENT FOR DETAILS). Assignors: CHARLES STARK DRAPER LABORATORY
Active legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R33/00—Arrangements or instruments for measuring magnetic variables
    • G01R33/20—Arrangements or instruments for measuring magnetic variables involving magnetic resonance
    • G01R33/44—Arrangements or instruments for measuring magnetic variables involving magnetic resonance using nuclear magnetic resonance [NMR]
    • G01R33/46—NMR spectroscopy
    • G01R33/4625—Processing of acquired signals, e.g. elimination of phase errors, baseline fitting, chemometric analysis

Definitions

  • MRS magnetic resonance spectroscopy
  • MRI magnetic resonance imaging
  • NMR nuclear magnetic resonance
  • the nuclei of these atomic isotopes such as H-atoms (most abundant isotope of hydrogen), carbon-13, oxygen-17, sodium-23, and phosphorus-31, are aligned in their lowest stable quantum states in the presence of a magnet.
  • RF radio frequency
  • NMR spectroscopy is a record of relative numbers of nuclei, which hop to the higher quantum state, versus frequency. The frequency will vary depending on the atom and the functional group (neighboring atoms) of the molecule to which the atom is attached.
  • NMR is ideal for imaging living tissue. Tissue contains an abundance of H-atoms, in the form of H 2 O and other organic molecules, and smaller but still significant amounts of sodium-23, phosphorous-31, etc. NMR is safer than X-ray based imaging (such as computerized axial tomography (CAT) scans) as the radio frequencies are much less energetic than X-rays. NMR-based MRI yields high resolution 2-D image slices from various portions of human body, which are now routinely used to identify injured tissue and tumors.
  • CAT computerized axial tomography
  • MRS In-vivo MRS is a specialized version of MRI, which, instead of 2-D image slices, is represented as a spectral curve of signal strength versus chemical shift in parts per million (ppm), defined as RF value divided by a carrier frequency in MHz. MRS is used, for example, to gauge metabolic changes in brain tumors, in areas of the brain affected by stroke and seizure, etc., by using isotope and functional group abundances of metabolites.
  • ppm parts per million
  • MRS scans can cover 1000 times larger volumes (1-10 cm 3 ) than MRI scans (1-10 mm 3 ). Because of the larger volume, MRS data includes H-atom (proton) peaks not just from water but also from other molecules and functional groups, unlike MRI which just includes water. According to Blüml, modern MRS scanners have built-in software to analyze raw MRS data by using approved scripts which apply line broadening, Fourier transform and phasing, and the final result is delivered to the physician/radiologist.
  • MRS signal Although analysis of MRS signal provides valuable information for medical diagnosis and addressing basic research questions, two of the many challenges of working with MRS data are the level of noise and artifacts than can arise in the raw signals and the disparity across data sets acquired from different scanners.
  • Raw data may have variable resolutions, i.e. points/ppm, variability in manufacturer formats and slightly different acquisition conditions. Some software packages can also combine different scans. Some of these software packages and algorithms are: recent INSPECTOR package associated with MATLAB (Columbia Technology Ventures, # cu17130, 2017, by Juchem); jMRUI2XML java package (Mocioiu et al., BMC Bioinformatics, 16:378, 2015) and LC model (Provencher, “LCModell & LCMgui User's Manual”; Provencher, “Estimation of metabolite concentrations from localized in vivo proton NMR spectra”, Magn Reson Med., 30(6):672-9, 1993); and, finally, Mirzaalian, et al., “Harmonizing Diffusion MRI Data Across Multiple Sites and Scanners”, Med Image Comput Comput Assist Interv. 9349: 12-19, 2015 (published
  • This invention concerns a method and a system for signal estimation and data harmonization for MRS (magnetic resonance spectroscopy) scans from one or more commercial scanners.
  • MRS magnetic resonance spectroscopy
  • MRS signals provide valuable information for medical diagnosis and addressing basic research questions.
  • Two of the many challenges of working with MRS data are the level of noise and artifacts than can arise in the raw signals and the disparity across data sets acquired from different devices. To address both of these issues, robust statistical methods based on functional analysis are applied to the raw signal.
  • This invention describes a method and a system for analysis of raw MRS data, in the form of signal strength versus chemical shift (ppm), from multiple scanners, including “signal estimation” from each raw data set, followed by cross-scanner “data harmonization” of results.
  • the final resulting MRS signals are consistent from one scanner to another, and are used for analysis by radiologists and other physicians.
  • the system and method typically include two consecutive processes: signal estimation and cross-scanner data harmonization.
  • the signal estimation process uses raw unprocessed MRS data which include signal strength as a function of chemical shift in ppm.
  • the raw data is specific to a control subject of a scanner.
  • Each raw dataset is processed to eliminate outlier spectra.
  • the remaining spectra are used generate an “ideal” MRS spectrum for the control subject.
  • the scanner outputs for a number of control subjects from one or more scanners are made compatible to the same from a single reference scanner.
  • Data harmonization describes the process of adjusting and normalizing the data acquired from one scanner to ensure that the amplitudes and frequencies of the peaks can be compared to a reference scanner. This harmonization process facilitates comparison of relative abundances of specific metabolites as measured by multiple scanners.
  • the embodiments in the invention use means and variances of corresponding MRS scans from a reference scanner and a second scanner to achieve data harmonization.
  • the MRS scans are obtained via signal estimation of raw MRS data for the same control subjects.
  • multiple scanners can be pair-wise harmonized with the same reference scanner.
  • mapping derived from quantitative comparison (using means and variances) of scanner outputs of a reference scanner to corresponding scans (same control subjects) from another scanner (target scanner).
  • mapping will be different for each target scanner even with the same reference scanner.
  • the invention features an MRS system. It comprises multiple MRS scanners producing MRS data of subjects. An analysis system then performs signal estimation to eliminate outlier spectra and a data harmonization process in which the spectra from control subjects from different scanners are made compatible.
  • the analysis system performs the harmonization process by adjusting and/or normalizing the data acquired from one scanner, by using another higher quality scanner as reference, to ensure that the amplitudes and frequencies of the peaks can be compared to the reference scanner.
  • the control subjects are scanned by each of the scanners.
  • the signal estimation is preferably repeated until certain threshold conditions are met.
  • Example conditions include a minimum percentage of raw data that must be included to compute the mean and a specified signal-to-noise (SNR) ratio.
  • SNR is the mean value of the spectrum divided by the standard deviation of the spectrum, performed pointwise on a function set and integrated over the range of 0-4 ppm.
  • the invention features a method for harmonizing MRS data.
  • the method comprises producing raw MRS data of control subjects and performing signal estimation to eliminate outlier spectra; and data harmonization in which the spectra from the control subjects from different scanners are made compatible.
  • FIG. 1A is a flow diagram showing the steps and flow of data for signal estimation which estimates MRS signal from raw data for a given scanner and control subject.
  • FIG. 1B is a system block diagram showing the steps and flow of data for data harmonization which harmonizes scans from scanners, such as those from a reference scanner and a target scanner.
  • FIG. 2A shows raw MRS spectra from a noisy collection showing relative abundances of H-atoms (signal strength) as a function of chemical shift (ppm).
  • the data is from a GE scanner (GE 3T 750w).
  • FIG. 2B shows signal to noise ratio (SNR) from 0-4 ppm for the entire collection of data in FIG. 2A , with all curves included (flat line) and for smaller sets of data where certain number of curves are dropped based on their deviation from the mean curve.
  • SNR signal to noise ratio
  • FIG. 3A shows raw MRS spectra from a higher quality collection (less noisy source) showing signal strength as a function of ppm.
  • the data is from a Siemens scanner (Siemens 3T Skyra).
  • FIG. 3B shows SNR from 0-4 ppm for the entire collection of data in FIG. 3A , and for smaller sets of data where certain number of curves are dropped based on their deviation from the mean curve.
  • the x-axis here denotes percentage of data included.
  • FIG. 4 shows individual MRS spectra, signal strength versus ppm, from the Siemens scanner.
  • the spectra are the result of signal estimation.
  • FIG. 5A shows individual MRS spectra, signal strength versus ppm, from the GE scanner.
  • the spectra are the result of signal estimation.
  • FIG. 5B shows the GE spectra of FIG. 5A after harmonizing them with respect to Siemens scans.
  • FIG. 6A shows MRS spectra from the Siemens scanner with 95% confidence bounds. The spectra are a subset of those shown in FIG. 4 .
  • FIG. 6B shows spectra similar to 6 A for the GE scanner. The spectra are a subset of those shown in FIG. 5A .
  • the term “and/or” includes any and all combinations of one or more of the associated listed items. Further, the singular forms and the articles “a”, “an” and “the” are intended to include the plural forms as well, unless expressly stated otherwise. It will be further understood that the terms: includes, comprises, including and/or comprising, when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. Further, it will be understood that when an element, including component or subsystem, is referred to and/or shown as being connected or coupled to another element, it can be directly connected or coupled to the other element or intervening elements may be present.
  • FIGS. 1A and 1B show the steps and flow for estimating MRS signals from raw data (signal estimation), and then harmonizing MRS scans (data harmonization) from two different scanners S 1 , S 2 according to the principles of the present invention.
  • One of the scanners, in this case S 1 is chosen as the reference scanner.
  • the other, S 2 is the target scanner.
  • the reference scanner should be the one that produces higher quality scans. Without that prior knowledge, the choice of reference is arbitrary. If more than two scanner results are being harmonized, the highest quality scanner among the lot of scanners should be the reference scanner. If that information is not known, any one scanner can be the reference scanner. The spectra from all other scanners are pair-wise harmonized with respect to the reference scanner.
  • FIG. 1A shows the signal estimation process flow 100 starting with raw MRS data 105 from a scanner.
  • the output 101 of the signal estimation process for a given scanner and a control subject is referred to as the MRS scan.
  • FIG. 1B shows results from harmonizing scans from two scanners S 1 , S 2 in the data harmonization step.
  • S 1 is chosen to be the reference scanner, the scans of which will be unchanged in the harmonization process.
  • the scans from S 2 will be changed (i.e., harmonized).
  • step 110 one starts with raw MRS data s i (z) 105 from one control subject from a scanner and then a median spectrum g(z) is selected.
  • a median spectrum g(z) is selected.
  • the subscript i denotes an individual spectra in the raw data collection;
  • s and g denote signal strength; and
  • z is in ppm.
  • Deviations of each signal from the mean are computed and signals with large deviations are dropped in step 120 .
  • a mean (as a function of ppm, z) is calculated
  • the median spectrum g(z) could be in a tabular form of strength defined on a discrete set of z-values.
  • the g-values for the discrete z-values could be the median or average values of s i at the corresponding z-value.
  • g(z) could be one of the s i curves which “snakes” through the middle of the s i collection.
  • Results from step 130 can be fed back into step 120 to repeat (iterate) the cycle until specified threshold condition(s) are met.
  • One threshold could be, for example, a minimum percentage of raw data that must be included to compute the mean.
  • Another threshold could be a specified signal-to-noise (SNR) ratio.
  • Box 101 contains the final mean of raw MRS data. This is the estimated scanner signal from the signal estimated step. This signal will be referred to as the MRS scan for the control subject and the scanner under consideration.
  • an “MRS scan” is the output 101 of the signal estimation process that starts with raw MRS data 105 as input.
  • the data harmonization process harmonizes MRS scans from target scanner with respect to those from a reference scanner.
  • S 1 will be the reference scanner.
  • the reference MRS scan for a control subject (control subject 1) is shown in box 101 - 1 - 1 using the notation 101 - m - n where m denotes the scanner, n denotes the control subject, and as shown in FIG. 1A, 101 denotes the output of the signal estimation process.
  • 101 - 1 - 2 and 101 - 1 - 3 are S 1 scans for control subjects 2 and 3, respectively.
  • the raw MRS data from S 1 are 105 - 1 - 1 (for S 1 , control subject 1), 105 - 1 - 2 (S 1 , control subject 2) and 105 - 1 - 3 (S 1 , control subject 3).
  • the notation is 105 - m - n , where 105 denotes raw data ( FIG. 1A ) and m and n denote scanner and control subject, respectively.
  • Each raw MRS data set 105 undergoes the signal estimation process 100 and produces output 101 , which is referred to as MRS scan.
  • the right side of FIG. 1B uses scans from the target sensor, S 2 , whose scans are 101 - 2 - 1 (S 2 , control subject 1, using raw MRS data 105 - 2 - 1 ), 101 - 2 - 2 (S 2 , control subject 2, using raw MRS data 105 - 2 - 2 ), and 101 - 2 - 3 (S 2 , control subject 3, using raw MRS data 105 - 2 - 3 ).
  • the control subjects are the same as those for reference scanner S 1 .
  • the notation used here are 101 - m - n and 105 - m - n .
  • m is 2 for the sensor (S 2 ) and n refers to the control subjects 1, 2 and 3.
  • 101 refers to output (MRS scan) of signal estimation process 100 and 105 refers to input (raw MRS data) to signal estimation 100 .
  • the scans from S 1 ( 101 - 1 - 1 , 101 - 1 - 2 and 101 - 1 - 3 ) and S 2 ( 101 - 2 - 1 , 102 - 2 - 2 and 101 - 2 - 3 ) are fed into the data harmonization engine 220 .
  • First mean and variances (which are first and second moments, respectively) for each scanner are computed in steps 211 (S 1 ) and 212 (S 2 ).
  • step 211 computes mean and variance using the scans from S 1 for control subjects 1 ( 101 - 1 - 1 ), control subject 2 ( 101 - 1 - 2 ) and control subject 3 ( 101 - 1 - 3 ).
  • step 212 uses target scanner S 2 scans for control subjects 1 ( 101 - 2 - 1 ), 2 ( 101 - 2 - 2 ) and 3 ( 101 - 2 - 3 ) to compute mean and variance.
  • Data harmonization process then uses the mean and variance of S 1 and S 2 to produce an empirical mapping 222 for data harmonization, which will transform S 2 scans 101 - 2 - 1 , 101 - 2 - 1 and 101 - 2 - 3 (see box 202 ) into harmonized scans 101 - 2 - 1 H, 101 - 2 - 2 H and 101 - 2 - 3 H (see box 202 H) so than they can be more easily compared with reference scans from S 1 , 101 - 1 - 1 , 101 - 1 - 2 and 101 - 1 - 3 .
  • the empirical mapping 222 is developed, it is used on original sensor 2 S 2 scans 101 - 2 - 1 , 102 - 2 - 2 and 102 - 3 , for control subjects 1, 2 and 3, to obtain three harmonized scans 101 - 2 - 1 H, 101 - 2 - 2 H and 101 - 2 - 3 H for the corresponding subjects.
  • S 1 scans 101 - 1 - 1 , 101 - 1 - 2 and 101 - 1 - 3 , remain unchanged.
  • the motivation behind harmonization is that the harmonized scans of S 2 are easier to compare against reference S 1 scans than un-harmonized scans.
  • n should ideally be much higher than 3, however it could be as low as 2.
  • additional scanners can be harmonized with respect to that reference scanner in a pair-wise manner. Ideally, if more than two scanners are involved, the reference scanner should be the same for all pair-wise harmonization processes. The reference scanner should be the one which produces most ideal scans. If this information is not available, the choice of reference scanner is arbitrary.
  • FIG. 2A shows a plot of a collection of raw MRS data for the GE scanner for a control subject.
  • the data are in the form of signal strength versus ppm.
  • FIG. 2B shows a plot of signal to noise ratio (SNR) from 0-4 ppm as a function of percentage of raw data included in computing the SNR ratio.
  • SNR signal to noise ratio
  • a 50% trim threshold refers to 50 percent of data with largest deviations being dropped from SNR computation.
  • the flat line is the SNR value (a little under 6) when all data are used to compute SNR.
  • SNR value is higher (about 9) when only 15% of lowest deviation curves are used to compute SNR. When more and more data are retained, SNR approaches the flat line for all data. Note that the deviations are calculated using the equation shown above.
  • the SNR results plotted are the result of iteration over eliminating outlier signals, re-computing the mean over the remaining signals and then computing SNR. This results in better SNR than yielded by the scanners which typically use all of the raw data without eliminating any outliers.
  • FIGS. 3A and 3B show similar plots as in FIG. 2A and FIG. 2B , but for the Siemens scanner.
  • FIG. 3A data has less of a spread than FIG. 2A data. Therefore, SNR does not show improvement unless more and more data are rejected.
  • FIG. 3B shows that unless more than 80% of data are rejected (i.e., only 20% or less are retained for SNR computation), SNR does not improve from the flat line value of about 11.4 for all data.
  • FIG. 4 shows individual MRS spectra from the Siemens scanner. Each scan is the result of signal estimation process 100 . In other words, each spectrum in FIG. 4 is “distilled” from a separate collection of a raw MRS data set. The peaks correspond to locations of H-atoms in various functional groups of molecules.
  • FIG. 5A shows individual MRS spectra from the GE scanner. Each scan is the result of signal estimation process 100 ; i.e., each spectrum in FIG. 5A is “distilled” from a separate collection of raw MRS data. The peaks correspond to locations of H-atoms in various functional groups of molecules.
  • FIG. 5B shows the results of harmonizing the GE scans with respect to Siemens.
  • the harmonization technique uses results of signal estimation method 100 to estimate the signals from multiple control subjects from each scanner.
  • the harmonization then uses an empirical mapping using variances and means 211 and 212 of MRS scans as described earlier. The goal is to insure that signals acquired from different scanners are comparable. Thus, scans of the same control subjects would exhibit the same statistical properties.
  • the GE spectra are qualitatively similar to the Siemens spectra of FIG. 4 .
  • the unaltered Siemens spectra were used as reference spectra.
  • FIG. 6A shows the Siemens scans from FIG. 4 , but the spectra which lie beyond the 95% confidence level, i.e., outside ⁇ 1.96 ⁇ standard deviation assuming Gaussian (normal) distribution of data, are omitted from the plot.
  • FIG. 6B shows the GE scans from FIG. 5A , but the spectra which lie beyond the 95% confidence level are omitted from the plot.
  • the Siemens plots are closer and more similar to each other than GE plots. This is likely a reflection of the dynamic nature of the sample (control subject) and instrument settings than the scanners themselves.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Condensed Matter Physics & Semiconductors (AREA)
  • General Physics & Mathematics (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)
  • Other Investigation Or Analysis Of Materials By Electrical Means (AREA)
US16/324,193 2016-08-10 2017-08-10 Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS) Active 2037-11-26 US11249158B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/324,193 US11249158B2 (en) 2016-08-10 2017-08-10 Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US201662372834P 2016-08-10 2016-08-10
US16/324,193 US11249158B2 (en) 2016-08-10 2017-08-10 Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS)
PCT/US2017/046219 WO2018031723A2 (en) 2016-08-10 2017-08-10 Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (mrs)

Publications (2)

Publication Number Publication Date
US20200191891A1 US20200191891A1 (en) 2020-06-18
US11249158B2 true US11249158B2 (en) 2022-02-15

Family

ID=60382583

Family Applications (1)

Application Number Title Priority Date Filing Date
US16/324,193 Active 2037-11-26 US11249158B2 (en) 2016-08-10 2017-08-10 Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS)

Country Status (4)

Country Link
US (1) US11249158B2 (de)
EP (1) EP3497461A2 (de)
AU (1) AU2017308905A1 (de)
WO (1) WO2018031723A2 (de)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020145424A1 (en) * 2001-04-04 2002-10-10 Nozomu Uetake Method of correcting resonance frequency variation and MRI apparatus
US20130131992A1 (en) * 2009-11-13 2013-05-23 The Government USA as Represented by the Secretary Dept. of Health and Human Services System for magnetic resonance spectroscopy of brain tissue for pattern-based diagnostics
US20160171670A1 (en) * 2014-12-16 2016-06-16 Siemens Aktiengesellschaft Image correction during magnetic resonance dependent on the reception profile of the reception antenna
US20160216352A1 (en) * 2013-08-30 2016-07-28 Koninklijke Philips N.V. Dixon magnetic resonance imaging

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020145424A1 (en) * 2001-04-04 2002-10-10 Nozomu Uetake Method of correcting resonance frequency variation and MRI apparatus
US20130131992A1 (en) * 2009-11-13 2013-05-23 The Government USA as Represented by the Secretary Dept. of Health and Human Services System for magnetic resonance spectroscopy of brain tissue for pattern-based diagnostics
US20160216352A1 (en) * 2013-08-30 2016-07-28 Koninklijke Philips N.V. Dixon magnetic resonance imaging
US20160171670A1 (en) * 2014-12-16 2016-06-16 Siemens Aktiengesellschaft Image correction during magnetic resonance dependent on the reception profile of the reception antenna

Non-Patent Citations (24)

* Cited by examiner, † Cited by third party
Title
"INSPECTOR: Magnetic Resonance Spectroscopy Software for Optimized Data Extraction," http://techventures.columbia.edu (2017).
Bluml, S., et al., "Magnetic Resonance Spectroscopy: Basics," MR Spectroscopy of Pediatric Brain Disorders, Springer-Verlag, 11-23 (2013).
Chard et al. "Reproducibility of In Vivo Metabolite Quantification With Proton Magnetic Resonance Spectroscopic Imaging," Journal of Magnetic Resonance Imaging, 15:219-225, 2003 (Year: 2003). *
Chard, D.T., et al., "Reproducibility of In Vivo Metabolite Quantification with Proton Magnetic Resonance Spectroscopic Imaging," Journa of Magnetic Resonance Imaging, 15: 219-225 (2002).
Fortin et al. "Removing inter-subject technical variability in magnetic resonance imaging studies," NeuroImage 132, 198-212, 2016 (Year: 2016). *
Gubian, M., et al., "Using Functional Data Analysis for investigating multidimensional dynamic phonetic contrasts," Journal of Phonetics, 49: 16-40 (2015).
Hall, P., et al., "Two-sample Tests in Functional Data Analysis Starting from Discrete Data," Statistica Sinica, 17: 1511-1531 (2007).
International Preliminary Report on Patentability, dated Feb. 21, 2019, from International Application No. PCT/US2017/046219, filed on Aug. 10, 2017. 13 pages.
International Search Report and Written Opinion, dated Mar. 29, 2018, from International Application No. PCT/US2017/046219, filed on Aug. 8, 2017. 20 pages.
Irvine, J. M., et al., "Robust Functional Data Analysis Approach to Data Harmoization and Signal Estimation," International Society for Magnetic Resonance in Medicine: Workshop on MR Spectroscopy: From Current Best Practice to Latest Frontier, (2016).
James, G.M., et al., "Performing Hypothesis Tests on the Shape of Functional Data," Computational Statistics and Data Analysis, 50(1): 1-18 (2005).
Levitin, D.J. et al., "Introduction to Functional Data Analysis," Canadian Psychological Association, 48(3): 135-155 (2007).
Lin, P.D., et al., "Cleaning Data the Chauvenet Way," SESUG, Proc. South East SAS Users Gr., Hilton Head, 2007.
Mirzaalian, H., et al.,"Harmonizing Diffusion MRI Data Across Multiple Sites and Scanners," Med Image Comput Comput Assist Interv., 9349: 12-19 (2015).
Mocioiu, V., et al., "From Raw Data to Data-Analysis for Magnetic Resonance Spectroscopy—The Missing Link: jMRUI2XML," BMC, Bioinformatics, 16: 378, 2015.
Provencher, S., "LCModel & LCMgui User's Manual," 1-184 (2016).
Provencher, S.W. "Estimation of metabolite concentrations from localized in vivo proton NMR spectra," Magn. Reson. Med., 30(6):672-679 (1993).
Ramsay, J.O. & B.W.Silverman, Functional Data Analysis, 2nd Ed., New York: Springer (2005).
Siddique, J., et al., "Multiple imputation for harmonizing longitudinal non-commensurate measures in individual participant data meta-analysis," Statistics in Medicine, Published online in Wiley Online Library (2014).
Sun et al. "Histogram-based normalization technique on human brain magnetic resonance images from different acquisitions," BioMed Eng Online, 14:73; 215 (Year: 2015). *
Van de Bank, B.L., et al., "Multi-Center Reproducibility of Neurochemical Profiles in the Human Brain at 7T," NMR Biodmed. 28: 306-316 (2015).
Van de Blank et al., "Multi-center reproducibility of neurochemical profiles in the human brain at 7 Tesla," NMR Biomed. 28:306-315, 2015 (van de Blank). (Year: 2015). *
Wang, J.L., et al., "Review of functional data analysis," Annu. Rev. Statist., 1-41 (2015).
Zandt, HJA in 't, et al., "Common Processing of in vivo MR Spectra," NMR Biomed, 14: 224-232 (2001).

Also Published As

Publication number Publication date
WO2018031723A9 (en) 2019-03-28
WO2018031723A2 (en) 2018-02-15
US20200191891A1 (en) 2020-06-18
AU2017308905A1 (en) 2019-02-28
WO2018031723A3 (en) 2018-04-26
EP3497461A2 (de) 2019-06-19

Similar Documents

Publication Publication Date Title
Akçakaya et al. Scan‐specific robust artificial‐neural‐networks for k‐space interpolation (RAKI) reconstruction: database‐free deep learning for fast imaging
JP5349673B2 (ja) データ補正装置
US8692551B2 (en) Magnetic resonance imaging water-fat separation method
EP1830198B1 (de) Datenkorrekturgerät und Datenkorrekturverfahren
Chen et al. Tensor image enhancement and optimal multichannel receiver combination analyses for human hyperpolarized 13C MRSI
US10705168B2 (en) System and method for low rank approximation of high resolution MRF through dictionary fitting
EP3408679B1 (de) System und verfahren zur optimierten diffusionsgewichteten bildgebung
Girard et al. Magnetization transfer from inhomogeneously broadened lines (ihMT): improved imaging strategy for spinal cord applications
US20090093709A1 (en) Noise reduction system and methods for magnetic resonance imaging
US11131738B2 (en) Multiband, multishot magnetic resonance elastography
CN113808176B (zh) 基于装置成像坐标的mr图像配准方法、分析方法及其装置
US11125847B2 (en) System and method for out-of-view artifact suppression for magnetic resonance fingerprinting
Sundaram et al. Individual resting-state brain networks enabled by massive multivariate conditional mutual information
US10042024B2 (en) Magnetic resonance imaging method and apparatus
Clarke et al. Three‐dimensional, 2.5‐minute, 7T phosphorus magnetic resonance spectroscopic imaging of the human heart using concentric rings
Wang et al. Denoising magnetic resonance spectroscopy (MRS) data using stacked autoencoder for improving signal‐to‐noise ratio and speed of MRS
Christensen et al. A user independent denoising method for x‐nuclei MRI and MRS
US20240398253A1 (en) Physics-based algorithm to universally standardize routinely obtained clinical T2-weighted images
Sharif et al. Patient‐adaptive reconstruction and acquisition in dynamic imaging with sensitivity encoding (PARADISE)
CN116230239A (zh) 用于磁共振弥散加权成像的多参数后处理模型的构建方法
US10386435B2 (en) Methods and systems for fast auto-calibrated reconstruction with random projection in parallel MRI
WO2008126017A2 (en) Motion corrected multinuclear magnetic resonance imaging
US11249158B2 (en) Method and procedure for signal estimation and data harmonization for magnetic resonance spectroscopy (MRS)
Hansen et al. Fast Padé transform accelerated CSI for hyperpolarized MRS
US20220343565A1 (en) Reconstruction apparatus, method, and program

Legal Events

Date Code Title Description
FEPP Fee payment procedure

Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY

FEPP Fee payment procedure

Free format text: ENTITY STATUS SET TO SMALL (ORIGINAL EVENT CODE: SMAL); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

STPP Information on status: patent application and granting procedure in general

Free format text: FINAL REJECTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: ADVISORY ACTION MAILED

AS Assignment

Owner name: THE CHARLES STARK DRAPER LABORATORY, INC., MASSACHUSETTS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:IRVINE, JOHN M.;MARIANO, LAURA J.;REEL/FRAME:057527/0417

Effective date: 20210916

STPP Information on status: patent application and granting procedure in general

Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION

STPP Information on status: patent application and granting procedure in general

Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS

AS Assignment

Owner name: THE BRIGHAM AND WOMEN'S HOSPITAL, INC., MASSACHUSETTS

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:LIN, ALEXANDER P.;REEL/FRAME:058128/0900

Effective date: 20211116

STPP Information on status: patent application and granting procedure in general

Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT VERIFIED

STCF Information on status: patent grant

Free format text: PATENTED CASE

AS Assignment

Owner name: UNITED STATES GOVERNMENT, MARYLAND

Free format text: CONFIRMATORY LICENSE;ASSIGNOR:CHARLES STARK DRAPER LABORATORY;REEL/FRAME:069049/0666

Effective date: 20240606

Owner name: UNITED STATES GOVERNMENT, MARYLAND

Free format text: CONFIRMATORY LICENSE;ASSIGNOR:CHARLES STARK DRAPER LABORATORY;REEL/FRAME:069049/0641

Effective date: 20240606

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YR, SMALL ENTITY (ORIGINAL EVENT CODE: M2551); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY

Year of fee payment: 4